In my last year of University, I built a homebrew 6809 machine - this would have been early '80s. It had 4K of CMOS static RAM, a serial port (6850), a parallel port (6522) and was on one double sized Eurocard in a custom made box with a power supply I put together myself. There was a backplane for up to four cards and this was the first. It had a small 4x4 keypad and a small 8 digit 7 segment display. I programmed the first EPROM by hand (in an exercise book in hex code) and a friend burned a 2732 for me on his machine.
The next board I built had another 4K of CMOS RAM, a custom 2732 character generator ROM, a 6545 video controller chip and a UHF transceiver to send an image to a black and white TV. I had 40x24 video running from it, and replaced the 4x4 keyboard with a full ASCII keyboard.
The third board was a 64K dynamic RAM board of which 48K was visible. Since the 6809 doesn’t have any dynamic RAM refresh, I designed my own with a counter to present the row and column addresses on the alternate cycle of bus the 6809 has. To my amazement it worked first time!
The fourth and most ambitious board was a floppy disk controller based around two Western Digital chips. I got hold of a 2nd hand 5.25" full height floppy drive, bought FLEX 09 and changed the firmware to support it.
I had planned to redesign it to one Eurocard - in fact it was to be the thesis for my MSc - however I never completed it and lost interest as I got into PC’s and bought my first 486 machine.
As I wirewrapped the whole thing, the first board turned black as the wires oxidised and the machine stopped working. I remember thinking I should photograph the boards… but that was long before digital photography so I decided not to.
The 6809 is a CPU I never really got into back then - it was 6502 all the way for me. I suspect just a general lack of systems out there that actually used it so there was little interest, although where I was working/studying, we did get one 6809 system with a floppy drive that ran Pascal (slowly!) but by then we also had Aim 65’s and BBC Micros, so it sort of just gathered dust in a corner and was forgotten…
Was this a somewhat common thing to happen or was this due to cost savings measures on your side?
(I’ve never heard of this, but wirewrapping was before my time.)
Wirewrapping was typically used for prototypes, but also for one-off or small scale productions. I once designed a DMA interface for a customer, they would only ever need one… so I wire-wrapped it. But our company also (back in the seventies/eighties) made small run productions of specialised data receivers and formatters, some of them were wire-wrapped as part of the production. I’ve also seen wire-wrapped boards from other companies.
Not the best photo, but this is a 6502 board I made up in the early 80’s - oddly enough I made this after I’d helped design and make some 6502 systems on PCBs - however for me as a hobby user then it was easier to wire wrap it than make my one-off PCB…
Normally the wirewrap creates a sealed contact between the wires and the square edges of the posts. So even if the outside of the wire were to oxidize the contacts should still be good.
The wires are normally aluminum, I think. The oxide for that is not black, but tends towards white instead.
Wire wrap wires are normally silver-plated, and I assume the inner core is copper (although I don’t know that). Silver sulfide is black, and this accounts for the blackened surface of wire wrap pins and wires, and of many RF components and connectors (which are, or were historically, also often silver-plated). It is correct, though, that the actual contact between the wrap wire and the wrapping post should not corrode (at least quickly), because it is actually welded.
If the boards blackened rapidly, however, I suspect that whatever environment caused that to happen was problematic in other ways.
Thanks for the information about the wires! No wonder they were pretty expensive if they had silver in them. Since the sockets had long gold plated pins I was not surprised at how much they cost.
